How to Declare AI Usage in Your Projects
Why Declare AI Usage?
LLM-generated code looks plausible but often contains subtle bugs, unchecked assumptions, and confident-sounding nonsense. When you ship AI-generated code without disclosure:
- Reviewers waste time figuring out why code breaks
- Maintainers inherit technical debt they didn’t sign up for
- Trust erodes when others discover the truth later
The solution is being honest about when you use any tool, especially LLM’s.
Two Honest Paths
| Path | Description | Badge Needed? |
|---|---|---|
| Vibe Code | LLM does most of the thinking; you’re the delivery mechanism | Yes |
| Personal Responsibility | You review every line, test it, own it | No |
The litmus test: Can you explain your changes in a code review without first looking at the diff? If not, you’re vibe coding.
Use the Sladge Badge
Add this to your README to signal heavy LLM involvement:
Markdown:
[](https://sladge.net)HTML:
<a href="https://sladge.net">
<img src="https://sladge.net/badge.svg" alt="AI Slop Inside" />
</a>Add a Disclaimer Section
Add to your README after the badge:
## AI Usage
This project was built with significant LLM assistance. Expect:
- Rough edges and incomplete error handling
- Code that may need review before production use
- Features implemented for demonstration, not durability
Contributions welcome to harden this code for real-world use.Add an AI Policy File
Create AI_POLICY.md in your repo:
# AI Usage Policy
## This Project
- **Uses LLMs for**: [e.g., initial implementation, boilerplate, prototypes]
- **Human review**: [yes/no just be honest]
- **Production ready**: [yes/no]
## For Contributors
If submitting AI-assisted code:
1. Mark PRs with `[AI-assisted]` prefix
2. Document what you asked the LLM to do
3. Test thoroughly before submittingThe Point
If you let an LLM write your code and you didn’t verify it, you didn’t “use AI to be more productive.” You generated technical debt and passed the cleaning cost to someone else.
Transparency lets others decide whether to trust, review, or avoid your code; that’s respect.
Related
- sladge.net: The original inspiration
- Ghostty AI Policy: Example enterprise approach
Crepi il lupo! 🐺